update outputs path so that we can mount workspace to /workspace/data (…

…axolotl-ai-cloud#1623) * update outputs path so that we can mount workspace to /workspace/data * fix ln order
Y-IAB · May 15, 2024 · 4fde300 · 4fde300
1 parent 3319780
commit 4fde300
Show file tree

Hide file tree

Showing 70 changed files with 72 additions and 69 deletions.
diff --git a/docker/Dockerfile-cloud b/docker/Dockerfile-cloud
@@ -21,7 +21,9 @@ RUN apt install --yes --no-install-recommends openssh-server tmux && \
     printf "\n[[ -z \"\$TMUX\"  ]] && { tmux attach-session -t ssh_tmux || tmux new-session -s ssh_tmux; exit; }\n" >> ~/.bashrc && \
     printf "[ ! -z \"\$TERM\" -a -r /etc/motd ] && cat /etc/motd\n" >> ~/.bashrc && \
     chmod +x /workspace/axolotl/scripts/cloud-entrypoint.sh && \
-    chmod +x /root/cloud-entrypoint.sh
+    chmod +x /root/cloud-entrypoint.sh && \
+    mkdir -p /workspace/data/axolotl-artifacts && \
+    ln -sf /workspace/data/axolotl-artifacts /workspace/axolotl/outputs
 
 ENTRYPOINT ["/root/cloud-entrypoint.sh"]
 CMD ["sleep", "infinity"]
diff --git a/examples/cerebras/btlm-ft.yml b/examples/cerebras/btlm-ft.yml
@@ -38,7 +38,7 @@ wandb_watch:
 wandb_name:
 wandb_log_model:
 
-output_dir: btlm-out
+output_dir: ./outputs/btlm-out
 gradient_accumulation_steps: 1
 micro_batch_size: 1
 num_epochs: 1

diff --git a/examples/cerebras/qlora.yml b/examples/cerebras/qlora.yml
@@ -25,7 +25,7 @@ wandb_entity:
 wandb_watch:
 wandb_name:
 wandb_log_model:
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 batch_size: 4
 micro_batch_size: 4
 num_epochs: 2

diff --git a/examples/code-llama/13b/lora.yml b/examples/code-llama/13b/lora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/code-llama/13b/qlora.yml b/examples/code-llama/13b/qlora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/code-llama/34b/lora.yml b/examples/code-llama/34b/lora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/code-llama/34b/qlora.yml b/examples/code-llama/34b/qlora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/code-llama/7b/lora.yml b/examples/code-llama/7b/lora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/code-llama/7b/qlora.yml b/examples/code-llama/7b/qlora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/colab-notebooks/colab-axolotl-example.ipynb b/examples/colab-notebooks/colab-axolotl-example.ipynb
@@ -84,7 +84,7 @@
         "    type: alpaca\n",
         "dataset_prepared_path:\n",
         "val_set_size: 0.05\n",
-        "output_dir: ./qlora-out\n",
+        "output_dir: ./outputs/qlora-out\n",
         "\n",
         "adapter: qlora\n",
         "lora_model_dir:\n",

diff --git a/examples/dbrx/16bit-lora.yaml b/examples/dbrx/16bit-lora.yaml
@@ -10,7 +10,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.0
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 512
 sample_packing: false

diff --git a/examples/dbrx/8bit-lora.yaml b/examples/dbrx/8bit-lora.yaml
@@ -10,7 +10,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.0
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 512
 sample_packing: false

diff --git a/examples/dbrx/fft-ds-zero3.yaml b/examples/dbrx/fft-ds-zero3.yaml
@@ -10,7 +10,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.0
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 512
 sample_packing: false

diff --git a/examples/falcon/config-7b-lora.yml b/examples/falcon/config-7b-lora.yml
@@ -28,7 +28,7 @@ wandb_entity:
 wandb_watch:
 wandb_name:
 wandb_log_model:
-output_dir: ./falcon-7b
+output_dir: ./outputs/falcon-7b
 batch_size: 2
 micro_batch_size: 1
 num_epochs: 4

diff --git a/examples/falcon/config-7b-qlora.yml b/examples/falcon/config-7b-qlora.yml
@@ -42,7 +42,7 @@ wandb_entity:
 wandb_watch:
 wandb_name:
 wandb_log_model:
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 # QLoRA paper Table 9
 # - 16 for 7b & 13b

diff --git a/examples/falcon/config-7b.yml b/examples/falcon/config-7b.yml
@@ -28,7 +28,7 @@ wandb_entity:
 wandb_watch:
 wandb_name:
 wandb_log_model:
-output_dir: ./falcon-7b
+output_dir: ./outputs/falcon-7b
 batch_size: 2
 micro_batch_size: 1
 num_epochs: 4

diff --git a/examples/gemma/qlora.yml b/examples/gemma/qlora.yml
@@ -12,7 +12,7 @@ datasets:
   - path: mhenrichsen/alpaca_2k_test
     type: alpaca
 val_set_size: 0.1
-output_dir: ./out
+output_dir: ./outputs/out
 
 adapter: qlora
 lora_r: 32

diff --git a/examples/gptj/qlora.yml b/examples/gptj/qlora.yml
@@ -23,7 +23,7 @@ wandb_entity:
 wandb_watch:
 wandb_name:
 wandb_log_model:
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 gradient_accumulation_steps: 2
 micro_batch_size: 2
 num_epochs: 2

diff --git a/examples/jamba/qlora.yaml b/examples/jamba/qlora.yaml
@@ -10,7 +10,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.0
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 4096
 sample_packing: false

diff --git a/examples/jamba/qlora_deepspeed.yaml b/examples/jamba/qlora_deepspeed.yaml
@@ -10,7 +10,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.0
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 4096
 sample_packing: false

diff --git a/examples/jeopardy-bot/config.yml b/examples/jeopardy-bot/config.yml
@@ -21,7 +21,7 @@ wandb_entity:
 wandb_watch:
 wandb_name:
 wandb_log_model:
-output_dir: ./jeopardy-bot-7b
+output_dir: ./outputs/jeopardy-bot-7b
 gradient_accumulation_steps: 1
 micro_batch_size: 1
 num_epochs: 4

diff --git a/examples/llama-2/fft_optimized.yml b/examples/llama-2/fft_optimized.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.05
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/llama-2/gptq-lora.yml b/examples/llama-2/gptq-lora.yml
@@ -33,7 +33,7 @@ wandb_project:
 wandb_watch:
 wandb_name:
 wandb_log_model:
-output_dir: ./model-out
+output_dir: ./outputs/model-out
 gradient_accumulation_steps: 1
 micro_batch_size: 1
 num_epochs: 4

diff --git a/examples/llama-2/lisa.yml b/examples/llama-2/lisa.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.05
-output_dir: ./lisa-out
+output_dir: ./outputs/lisa-out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/llama-2/loftq.yml b/examples/llama-2/loftq.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/llama-2/lora.yml b/examples/llama-2/lora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/llama-2/qlora-fsdp.yml b/examples/llama-2/qlora-fsdp.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.05
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/llama-2/qlora.yml b/examples/llama-2/qlora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/llama-2/relora.yml b/examples/llama-2/relora.yml
@@ -12,7 +12,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./relora-out
+output_dir: ./outputs/relora-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/llama-3/fft-8b.yaml b/examples/llama-3/fft-8b.yaml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.05
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 8192
 sample_packing: true

diff --git a/examples/llama-3/lora-8b.yml b/examples/llama-3/lora-8b.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 
 sequence_len: 4096
 sample_packing: true

diff --git a/examples/llama-3/qlora-fsdp-70b.yaml b/examples/llama-3/qlora-fsdp-70b.yaml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.05
-output_dir: ./out/qlora-llama3-70b
+output_dir: ./outputs/out/qlora-llama3-70b
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/llama-3/qlora.yml b/examples/llama-3/qlora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/mamba/config.yml b/examples/mamba/config.yml
@@ -12,7 +12,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.0
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 2048
 sample_packing: false

diff --git a/examples/mistral/bigstral-ds-zero3.yaml b/examples/mistral/bigstral-ds-zero3.yaml
@@ -23,7 +23,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.05
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 2048
 sample_packing: true

diff --git a/examples/mistral/config.yml b/examples/mistral/config.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path:
 val_set_size: 0.05
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 8192
 sample_packing: true

diff --git a/examples/mistral/lora-mps.yml b/examples/mistral/lora-mps.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 eval_sample_packing: false
 
 adapter: lora

diff --git a/examples/mistral/lora.yml b/examples/mistral/lora.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.1
-output_dir: ./lora-out
+output_dir: ./outputs/lora-out
 
 adapter: lora
 lora_model_dir:

diff --git a/examples/mistral/mistral-qlora-fsdp.yml b/examples/mistral/mistral-qlora-fsdp.yml
@@ -12,7 +12,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.02
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 model_config:
   output_router_logits: true

diff --git a/examples/mistral/mistral-qlora-orpo.yml b/examples/mistral/mistral-qlora-orpo.yml
@@ -16,7 +16,7 @@ datasets:
     type: chat_template.argilla
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.1
-output_dir: ./mistral-qlora-orpo-out
+output_dir: ./outputs/mistral-qlora-orpo-out
 
 adapter: qlora
 lora_model_dir:

diff --git a/examples/mistral/mixtral-8x22b-qlora-fsdp.yml b/examples/mistral/mixtral-8x22b-qlora-fsdp.yml
@@ -11,7 +11,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.02
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 model_config:
   output_router_logits: true

diff --git a/examples/mistral/mixtral-qlora-fsdp.yml b/examples/mistral/mixtral-qlora-fsdp.yml
@@ -12,7 +12,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.02
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 model_config:
   output_router_logits: true

diff --git a/examples/mistral/mixtral.yml b/examples/mistral/mixtral.yml
@@ -12,7 +12,7 @@ datasets:
     type: alpaca
 dataset_prepared_path: last_run_prepared
 val_set_size: 0.0
-output_dir: ./qlora-out
+output_dir: ./outputs/qlora-out
 
 ## You can optionally freeze the entire model and unfreeze a subset of parameters
 unfrozen_parameters:

diff --git a/examples/mistral/mixtral_22.yml b/examples/mistral/mixtral_22.yml
@@ -21,7 +21,7 @@ model_config:
 datasets:
   - path: yahma/alpaca-cleaned
     type: alpaca
-output_dir: ./out
+output_dir: ./outputs/out
 
 sequence_len: 8000
 sample_packing: true